Most manual wheelchair users with hemiplegia use both the unaffected arm and leg to propel their wheelchairs. The objective of this study was to compare the wheelchair propulsion of subjects using the hemiplegic pattern (one arm and one leg) with subjects using two hands. In a case-controlled study in a kinesiologic laboratory, nine wheelchair users who used the hemiplegic pattern were compared with nine matched controls who used the two-handed pattern. Participants were tested for propelling and stopping the wheelchair, forward and backward, on a level surface and on a 5° incline. Video recording was used to assess deviation from the midline, foot slippage, the number of propulsive cycles, and the propelling velocity. Also, on the 5° incline, we noted the need for support when unlocking the wheel locks, instances of grabbing the side rail, or rollback between propulsions. The participants using the hemiplegic pattern when propelling up the incline deviated more to the hemiparetic side (P< 0.05), used more propulsive cycles per unit of distance (P< 0.01), were slower (P< 0.001), and used the side rail more often (P< 0.05). When propelling forward on level ground, the participants using the hemiplegic pattern were slower (P< 0.005). When stopping after moving backward down the incline, they were more likely to deviate to the unaffected side (P< 0.01). In conclusion, wheelchair users who use the hemiplegic pattern experience more difficulties than those using two hands, some of which may be amenable to improvements in wheelchair prescription and training.